On Designing of Modified Exponentially Weighted Moving Average Control Chart based on Sign Rank for Zero-Inflated Data

摘要

The performance of parametric and nonparametric control charts that are carefully created for zero-inflated count data—which are commonly modeled using zero-inflated Poisson (ZIP) and zero-inflated binomial (ZIB) distributions—is thoroughly assessed in this study. The evaluation of the three primary control chart performance metrics—average run length (ARL), median run length (MRL), and standard deviation run length (SRL)—forms the basis of the comparative analysis. The performance evaluation’s new results clearly show that nonparametric control charts are more effective at identifying shifts in the probability of small-scale inflation. On the other hand, changes in large-scale inflation are better detected by parametric control charts. These empirical observations are further substantiated through the applications of real-world data, which serves as a robust case study for assessing the effectiveness of both types of control charts in practical settings.